The place of place in creole genesis

 THE PLACE OF PLACE IN CREOLE GENESIS

Hildo Honório do Couto

University of Brasília, Brazil
Email: hildodocouto@gmail.com

1. Introduction*

It is perfectly possible that in the upcoming century the notion of space may lose much of its present-day importance for most cultures. The notion of territory as a symbol for the community of speakers of the several languages in the world may disappear through the increasing globalization process of culture and communication. In this case, the  close contact of the speakers with one another may become less relevant than it is today. In fact, even at present it is already possible to preview this in what we could call “Internet communication”, among several others. When we receive a message through the Internet, it does not matter whether it comes from Japan, Germany, Brazil, the nearest corner or from our neighbor next door. Frequently we receive a message from an unknown person, respond to it, and engage in  “conversation” with him/her, without knowing where s/he lives. In short, it is likely that we will live in a common world of a “global community”, as Marshall MacLuhan had foreseen more than 30 years ago -- he used the term “global village” (see MacLuhan & Fiore 1967). It is possible that in a still more distant future we may have even an “interplanetary”, or “intergalactic, community”.

Despite this forecasting for the future, in the emergence of most known pidgins and creoles space has played a very important role. Therefore, the  aim of this paper is to explore the relationship between language and the territory in which its speakers live, illustrating with what happened in Brazil, especially in the maroon community known as Palmares, and in Guinea-Bissau.

In order to assess this relationship, it is important to first present the framework in which it will be discussed. I depart from a very simple and obvious concept, namely that of community. It will be understood as consisting of a population (P) living in a territory (T) and unified by a language (L). The second concept to be discussed in a relatively detailed way is the second component of community, i.e., territory (T). It will be considered in close relation to L. In this connection, I will defend the thesis that a creole language begins to emerge when a “medium of interethnic communication” (see Baker 1990, 1997) becomes the primary language (L) of a population (P) living in a certain territory (T).

I think that this is a good starting point for a revision of some creolistics concepts that have proven to be problematic such as that of “pidgin” and the consequent one of creole as its “nativization”. In fact, some creolists are already revising them. Among them we could mention Mufewne (1997), Thomason (1997) and Singler (1992), not to mention those for whom “nativization of a pidgin” has never played any role in creole grammar formation (see Bollée 1977, Valdman 1977, Chaudenson 1989, Alleyne 1971).    

2. Community

Probably nobody would disagree with the assertion that in order for a language (L) to exist there must be a population (P), which speaks it, living in a certain territory (T). This may be schematically represented as in the figure below. Hence, throughout this paper, community (C ) is understood as represented in the formula C = PTL.

L

/    \

P-----T

Model of Community

As a mater of fact, the main concern of us creolists is the upper angle of the above triangle, namely, creole languages (L). However, some creolists have been studying the importance of its left bottom angle (P) as well. It is the case, for instance, of Philip Baker and, more recently, Mikael Parkvall. The right bottom angle (T) has not been an object of investigation by most researchers. Therefore, my main objective here is to emphasize its importance in creole grammar formation.

Not all aggregations of individuals thrown together in the same place make up a community. In order for there to be a community as defined here, the aggregation of individuals must have a series of interests in common, thus justifying the term community. In fact, both etymologically and in common sense, “community” means “something held in common”. One of the most important aspect of these commonalities is language. Therefore, L is representing all the remaining commonalities in the above model.

We may go further and define society (S). It is the aggregation of individuals (P) unified by L, therefore,  S = PL. If we consider the aggregation of individuals (I₁, I₂, ...., I_n) without some interests in common - without L -  we have what is called a kinetic aggregation, i.e., PT. As to the relation T-L, it will be the subject of the whole of section 3, besides the fact that it pervades this essay throughout.

At face value, this concept of community is obvious and, consequently, does not deserve being studied at all. However, I think that if we consider it as a point of departure, and not of arrival, it can be  a good framework for discussing the formation of creole grammar, i.e., of creole language. First of all, we can formulate precisely the idea of language contact. As a matter of fact, languages alone do not get into contact. What comes into contact are peoples -- let’s say P₁ and P₂ -- together with their respective languages, in this instance L₁ and L₂. Therefore, in a contact situation what we have is PL₁ getting into contact with PL₂ (see section 4). Second, the model suggests in a straightforward way that the territory (T) plays an important role in the emergence and existence of languages, by the simple fact that it is part of the same whole as L. As early as the end of the last century, Lewis Morgan suggested a similar model. Speaking of the Iroquois Indians, he said that “each tribe was individualized by a name, by a separate dialect, by a supreme government, and by the possession of a territory which it occupied and defended as its own” (Morgan 1878: 102; see also 112-121).

It is no coincidence that the term communication has the same root as community. If  individual I₁ and individual I₂ do not belong to the same population (P), hence to the same community, they do not share the same means of communication -- including language. Therefore, they cannot understand one another. However, if they do belong to the same P, i.e., to the same community, they share a lot of experiences, which they may communicate to one another. This is the case even if they are dislocated, namely, when they are temporarily or permanently in another community. If an individual I₃, belonging to another community -- thus  having a different language and different experiences -- wants to communicate with individual I₁ or with individual I₂, s/he must first learn their L.

As can be seen, the well-known model of communication seen below is also assumed here. According to this model, when one individual (S) says something (M) to another (H), M will be understood by H only if formulated in an L s/he shares with S.

L

/    \

/        \

S-->M-->H

Model of communication

(L = language; S = sender; M = message; H = hearer)

This is another very simple and obvious idea. However, besides being closely connected with that of community, it is a way to gauge whether an aggregate of people has a common L or not, from a synchronic point of view -- whenever S_i menages to communicate linguistically with H_i. Diachronically, it allows us to show that a new language emerges out of  several interactions between S and H of mutually unintelligible languages. However, this point will not be discussed here.

3. Language and territory

The recognition of the relationship between language and space is not new. One of its first and most important manifestations may be seen in the dialectological school called dialect geography. The methodological tools used by dialect geographers are largely taken from the science of geography proper. Among them we could mention “linguistic atlas”, “stratigraphy”, “isogloss” (as compared to isotherm), and so on. In this case, the relative proximity or distance from a certain point in space generally leads to a linguistic proximity or distance from the language of that point. Thus, dialects geographically more distant from the prestige center are more likely to be more different from the standard language than those closer to it. In this connection Lewis Morgan says that “separation in place, and distance between their settlements, had long before their [groups of Iroquois - HHC) discovery resulted in the formation of dialects, and in tribal independence” (1978: 107). Elsewhere he says that “separation of the people in area was followed in time by variation in speech” (p. 104).

Among common people, the relationship tetween L and T may manifest itself in several ways. In Germany, I heard several times the question: “Sprechen Sie Brasilianisch?” (do you speak Brazilian?). Further, I have seen several French books with the explanation: “Traduit de l’américain par.....”. In other words, “Brazilian” should be the language of Brazil, and “américain” the language of America, on the same foot as Italian is the language of Italy, and Portuguese the language of Portugal, as well as English is the language of England. Finally, at least in Brazil it is usual among common people to call certain individuals by their town or state of origin: “o baiano” (the Bahian, i.e., someone from the state of Bahia), “o Ceará” (the Ceará, i.e., someone from the state of Ceará), etc.

Coming back to scientific investigation, we can see that the relationship between space and language or, more generally, between  space and behavior is dealt with by several disciplines. Putting aside philosophy and physics, which deal with space in a very abstract sense, we could begin by mentioning ethology (see Eibl-Eibesfeldt 1968), which deals with animal behavior. Here space appears under the notion of territoriality. According to Thomas A. Sebeok, “territoriality refers to a variety of behavior patterns associated with the active defense of a certain site by the animal” (Sebeok 1972: 172). Semioticians, or more specifically, zoosemioticians, generally study the use of space as representation and/or communication under the name of proxemics, which was born in the context of anthropology, meaning “the study of man’s perception and use of space” (Hall 1968: 83). In semiotics it is defined in roughly the same way, viz., “proxemics, the study of man’s differential perception and use of space and time, is a complex of activities and their derivatives known to ethologists, since 1920, as territoriality” (Sebeok 1972: 166). Labrie (1996) presents an example of application of the concept of territoriality to language. Although he departs from a different perspective than mine, he acknowledges that “la notion de territoire demeure étroitement lié à celle de la langue” (p. 217).

Ralph B. Taylor (1988) gives a good example of a book-length treatment of the relationship between behavior and space (Taylor 1988: 117). Although he is a bit eclectic and up to a certain point impressionistic, as a general rule he follows what he calls “ecological psychology’s behavior setting theory”. That is, his main theoretical framework is psychology, or ecological psychology. At the outset he says that his “human territorial functioning” “applies largely to small groups, and the individuals in those groups, and is limited largely to small-scale, delimited space” (p. 1). As creolists well know, creole-speaking regions are rarely large countries. On the contrary, they are mostly islands or, at best, “continental islands” or “Sprachinsel” (Wiesinger 1973), as in the case of fort creoles.

Architecture is another case in which the relationship between social structure (culture, behavior) and space is of paramount importance. Donald Preziosi, for instance, says that the principles underlying the organization of architectonic space are the same underlying language in the strict sense. He uses terms such as “built environment”, “environmental structure”, and “architecture” to refer to this kind of space (see Preziosi 1979).

In the domain of linguistic investigation, some authors have tackled the relationship between L and T under the name of language ecology. According to Einar Haugen, “language ecology may be defined as the study of interactions between any given language and its environment”. Then he adds that “the true environment of a language is the society that uses it as one of its codes”. However, like Sapir (see below), he admits that there are both “social and natural environment”  (Haugen 1972: 325; see also Haugen 1979). He does not refer to space explicitly. However, in the latter essay he deals with a small community, located in a distant island, viz., Faroe. Another sociolinguist, William Mackey, takes up Haugen’s idea and applies it to both language contact, one of the main subject matters of creolistics, and to language shift (Mackey 1979, 1980). The idea of space (territory) pops up explicitly at least when he speaks of “geolinguistic shift” (1980: 36). However, it is implicit throughout his two essays .

Long before Haugen’s study, Edward Sapir had published his Language and environment (1912). Sapir makes in this essay the distinction between physical and social environment. Then he shows that most reflections of the former on language show up in the lexicon, rarely, if ever, in the grammar. About the relationship between both types of  environment and language he says: “It is the vocabulary of a language that most clearly reflects the physical and social environment of its speakers” (Sapir 1963: 90). He thinks that in “primitive” societies this relationship is closer than in a highly developed society. In the former, it would be possible to find certain reflections of the physical environment even in grammar.

Once I asked William Samarin whether it was possible for there to be a stabilized pidgin, i.e., an L,  without a community of its own. His answer was the following: “I would say yes, at least when we understand the words in a certain way. Take Sango, for example. It probably began to emerge as a pidgin in the year 1887-1888. Probably in less than twenty years it was a stable pidgin. Certainly it was stable in the 1950s when I learned it. Up until the 1960s there were very few people for whom it was the first language. In Bangui now there are thousands, and almost 50% of the preschool children I studied recently were native speakers of Sango. Earlier in this century was there a 'speech community'? If we use RULES in the sociolinguistic sense, we'd have to say yes. Early in the 1950s, when I went to Oubangui-Chari, Sango had more prestige than an ethnic language, although less than French. (See my paper in Bright's volume of 1966.) In many Protestant churches pastors would preach in Sango even though they were preaching in a village of Gbaya or Kaba or Tali people. These pastors were educated in Sango, and they felt apparently that Sango was the religious language. Such attitudes contribute to the creation of a COMMUNITY” (Samarin, p.c., through the Internet).

Although Samarin uses the word “community” in a different sense than the one attributed to it here, we can see that his answer is in the interrogative. Anyway, I do not think that the case of Sango would contradict the model presented in section 2, according to which language (L), is part of a whole called community (C), whose other component parts are population (P) and territory (T), i.e., C = PTL. What we may dispute is whether the T of the language we are talking about is an enclave in the T of the lexifier language or in that of the substrate languages or whether it is a third T, or the fourth possibility hinted at, but not discussed, in section 4 below. Before having its own T, the Sango probably consisted of the same universal simplification tendencies observable in what is called Russenorsk and Lingua Franca (see section 6).

Language exists mainly for interaction, for the exchange of information, between members of the community of which it is part. However, the mere existence of a system (lexicon + grammar) is not enough for the efficacy of this interaction. Shared experience is also needed. If individuals do not live together in a common territory -- in the same community -- they do not have the necessary shared experience, or not enough shared experience, to make effective communication possible.

Both non-verbal and verbal interaction -- communication -- are the beginning of the process of formation of any creole and, in the end, of any language. Therefore, it  was very common for slaves speaking several different languages to begin to come to a common denominator already during the Middle Massage. If they spoke different dialects of the same language, there was a tendency to a koineization on board. Even sailors tended to create a sort of “ship English” (see Hancock 1986: 85). In short, even in this case we have a sort of mini-community, in a mini-territory -- the ship -- that made the existence of an intragroup means of communication necessary (see also Rougé 1988: 9).

To my knowledge, there is no known case of a creole language emerging among, and being used by, a nomad people. At most, they might use a pidgin or contact language when dealing with the peoples they get in touch with. The existence of a fixed territory seems to be unavoidable for the emergence of creole languages, and of languages in general. However, the relationship between L and T is not a mechanically determined one, as in the well-known theory of climates of Montesquieu (see Montesquieu 1899: 221-234). Rather, it could be as suggested by Sapir, according to whom the influence of the physical environment on L is filtered by the social environment.

According to the model of community seen in the previous section, T is a necessary condition, albeit not a sufficient one, for the emergence and existence of a language specific to that community. From a historical point of view, T precedes P, which precedes L. In other words, first there must be a territory where a group of people come together. One of the first things they must somehow agree upon is what part of T will be occupied by whom. A second one could be a name for T itself.  

4. Language contact

When we talk about language contact, it must be clear that what comes into contact are not languages properly, but people (P) and their languages (L), i.e., PL. Thus, in the case of the contact-induced language change that gives birth to creole languages, we generally have the contact of a dominant people and their language (PL₁), with the slave and/or dominated peoples and their languages (PL₂, PL₃, ...., Pl_n) (see diagram of the first part of section 5). As diagramed below, the contact may take place (a) in the territory of (PL₁), i.e., in T₁, (b) in the T of one of  (PL₂, PL₃, ...., Pl_n) or in a territory common to them, i.e., in T₂, or (c) in a third territory, which is neither (PL₁)’s nor (PL₂, PL₃, ...., Pl_n)’s, i.e., in T₃. For short, let me refer to (PL₁) as LL (lexifier language), and to (PL₂, PL₃, ...., Pl_n) as SL (substrate languages).  

(a)                                           (b)                   (c)                  

LL + SL                                  LL + SL          LL +  SL

 |     /                                           \     |                 \   /  

 T₁                                                   T₂                T₃                          

Situation (a) would hardly give birth to a creole language. Some examples of this situation are the case of the Turkish, Italian and Yugoslavian immigrants in Germany as well as the Spanish-speaking immigrants in Miami and New York. Even when SL speakers manage to keep at least part of their language, what we generally have is a slow but  increasing acquisition of the  language of the host community along generations. One interesting case of language used in such a situation is Romani. In many places only part of the original vocabulary survives, whereas the grammar is wholly that of the surrounding language. Elsewhere I called this situation anticreole (Couto to appear). Anticreole is part of a larger phenomenon which Bakker dubbed intertwined languages. Media Lengua, and Chamorro are also instances of this situation (Bakker & Muysken 1996). In other words, they are mixed languages whose lexicon comes from one source but whose grammar comes from another source. As a rule, intertwined languages result from a relexification of LS by LL. Anticreole is their subset which results from a regrammaticalization of SL by LL, besides other characteristics such as resistance to assimilation, and language death.

Situation (b) is the case of the so-called fort creoles, like Guinea-Bissau Portuguese Creole and Krio. In many cases, LL annihilates SL almost in its entirety. This is what happened during the colonization of America and Asia by European powers. In some countries some aboriginal people still survive, albeit highly acculturated. Even where the local ethnic language did not disappear, it is being influenced by the language of the ex-colonizer at an increasing pace. This is what happens in Paraguay, Bolivia, Ecuador and Mexico, among others. In a country like  New Zealand aboriginal languages pratically disappeared. There are several other instances of this situation that deserve attention on the part of creolists. 

The third possibility (c) is what happened in Hawaii and Mauritius. It constitutes what Bickerton (1988) called plantation creoles. However, it includes cases like Cape Verde as well, that was rather an entrepôt for the distribution of slaves to several European colonies. At any rate, in situation (c), representatives of both LL and SL are distant from their homeland, and this favors the emergence of a common means of communication.

A forth possibility might exist if the contact takes place in both T₁ and T₂. This could happen in a contact situation where LL and SL would live in contiguous territories. This would be an ideal scenario for the emergence of a “pidgin” as traditionally defined, i.e., a contact “language” without native speakers. Although this point will not be developed here, I think that it is worth being pursued.

At the periphery of creole and pidgin languages proper, there is a great deal of complex linguistic situations resulting from the contact of speakers of mutually unintelligible languages in a certain T. According to John Reinecke (1937), first we have the so-called Sprachinseln (Wiesinger 1973). Besides languages of nomad peoples like Shelta and  gypsies’ Angloromani, there are sedentary peoples’ Sparchinseln like the Wends of Lusatia. Other cases would be the “colonial dialects (unrestructured varieties of European languages in the New World, Australia, etc.), foreigners’ mixed speech (e.g. the English of foreigners in Hawaii), dying minor languages (e.g. New Jersey Dutch), ‘babu’ language (a school-taught foreign language that becomes a secondary language), and lingua francas that have not been restructured” (see Holm 1988: 41).  

Yet another consequence of language contact is what John Holm calls semicreole, i.e., languages that present both creole and noncreole features. Some of the examples he gives are Afrikaans, Rural Brazilian Portuguese, Reunionnais, among others (Holm 1994). To these, we could add Halliday’s (1976) antilanguages as well as several types of group and/or professional jargons, such as Pachuco and many others.

In summary, all the above language contact situations -- and similar situations -- have somehow to do with space. In fact, contact only takes place when representatives of different PLs meet at a certain place.

5. Creole genesis

The fundamental question in creolistics is the formation and transformation of creole grammar. However, we must bear in mind that the process of formation and transformation of creole grammar is not simply a structural one. As Ferdinand de Saussure said, the complete grammar of a language can be found only in the community of speakers as a whole, not in the individual, as Chomsky puts it. Thus, “in order to have a language, there must be community of speakers” (Saussure 1989: 77). The formation of the grammar is the result of several acts of communication between individuals living together in the same territory, for whatever reason. Talking of the relationship between language and speech, Saussure said that “historically, speech always takes precedence. How could we ever associate an idea with a verbal sound pattern, if we did not first of all grasp this association in an act of speech? Furthermore, it is by listening to others that we learn our native language. A language accumulates in our brain only as the result of countless experiences”. To reinforce this idea, he added that “the former [language] is at the same time the instrument and the product of the latter [speech]” (Saussure 1989: 19). Therefore,  we must take into consideration both the linguistic (or structural) process and the sociohistorical process of the formation and transformation of creole grammar (see Thomason & Kaufman 1988: 35, 212).

The “classic” model to explain the formation and transformation of creole grammar is synthesized by Derek Bickerton in the following way: “A pidgin is an auxiliary language that arises when speakers of several mutually unintelligible languages are in close contact; by definition, it has no native speakers. A creole comes into existence when children acquire a pidgin as their native language” (Bickerton 1984: 173). See also Hall (1966) and Hymes (1971).

However, as we will see in section 6 below, this model of creole formation presupposes pidgin, which is a problematic concept. What is more, the idea of nativization itself is also problematic. Fortunately, there is another model in creolistics literature to explain the emergence of a creole, according to which a creole language emerges when a jargon -- or whatever is used by peoples speaking mutually unintelligible languages are put together in a specific territory -- becomes the primary language of the new community. Instead of the triangular model pidgin > creole > decreolization, we have a model similar to a U-turn, as shown in the figure below (see the end of section 6 for more details).

---------------------> creolization

(PL₂, PL₃,..., PL_n)      
­¯ ® IAC > ISC  > creole grammar
(PL₁)             
<--------------------- decreolization

The top arrow turned to the right shows the gradual formation of creole grammar, i.e., creolization. The bottom arrow turned to the left indicates the process of transformation of creole grammar, i.e., decreolization, putting aside the fact that this concept has been disputed by many creolists. The creolization line starts from the contact of (PL₂, PL₃,..., PL_n) with (PL₁), indicated by the vertical arrows (­¯). The first arrow (­) indicates that it is generally members of (PL₁) who first address members of (PL₂, PL₃,..., PL_n). Members of the second group generally address members of the first in response to their solicitation (¯). At the very beginning, we have mere individual attempts at communication (IAC). Those IAC that are understood by the hearer tend to crystallize as individual strategies of communication (ISC) somehow similar to the fossilizations in  SLA. Almost all the examples of HPE given by Bickerton (1981, 1984) come under this case. If these ISC socialize to a certain point, a medium of interethnic communication may emerge (Baker 1990). This is the beginning of creole grammar. Here we can see why Saussure said that language (grammar) is a product of speech (acts of communication). Decreolization is a kind of return to (PL₁). What is important here is the simplicity of the model. It implies only the formation and transformation of creole grammar, here called creolization and decreolization, respectively.

5.1. Palmares

Brazil shows all the socio-historical prerequisites for the emergence of a creole language. Its population was formed by the mixture of three races and cultures, i. e., the European colonizers (Portuguese), Africans and several Amerindian ethnic groups. The two latter groups spoke several different and, in general mutually unintelligible, languages. However, there is no way of proving whether there was (not) a creole in the past, due to the absolute absence of historical documents (see Couto to appear) and, at least partially, to the continental size of the country. Creoles tend to emerge in small regions such as the islands of the Caribbean and of the Indian and Pacific Ocean. Therefore, I will briefly discuss a maroon community, called Palmares, about which we have some documentation that points to the strongly possible existence of a creole.
Palmares existed approximately from 1630 up to 1697 in what is today the state of Alagoas, in  northeastern Brazil. From 1630 to 1644 the Dutch invasion of the region occurred . From 1644 to 1650 we have the period of restoration of Portuguese rule. It was located at a barely accessible place, near what is called Serra da Barriga (Belly’s Hill), and was blocked off with barricades against intruders.
The demographic composition of Palmares was the most varied possible. It consisted basically of slaves who escaped from slavery, many of them of Bantu origin. Some were from the Coast of Guinea. Among them there were also a few mulattoes and creoles as well as some Amerindians (Silva 1988: 23, 29-30). It is possible that even some whites could be found among them because the Palmareans captured and enslaved people to help them build their community. Palmares may have had from 16 to 20 thousand inhabitants.
The main aim of the Palmareans was to build a society of their own and free from slavery. Therefore, they had rigid laws, and breaking of these laws meant punishment by death. They were governed by a “king”, called Zumbi, one of the most famous characters in the history of Brazilian black community.
As we can see, all the socio-historical ingredients for the formation of a creole were present. The majority of the inhabitants of Palmares came from the sugar plantation of nearby Pernambuco. Therefore it is possible that they already knew some sort of pidginized/creolized Portuguese. On the other hand, it is highly improbable that they spoke some African language among themselves due to their varied origins, although it is possible that some kind of African lingua franca may have existed in the region outside Palmares’ territory (Rodrigues 1945, Rodrigues 1983). Their intention was to form a community different from the surrounding Brazilian one, which was their enemy. Therefore, they isolated their territory.
The documents of the time are evasive in regard to how the Palmareans communicated among themselves. They describe the demography, the topography, the architecture, the economy and even some lifestyles of the Palmares community. However, they are wholly silent in relation to the language(s) that was(were) used locally. It seems that the theme “language” was a kind of taboo. The chroniclers of the time were probably afraid of recognizing some kind of state in Palmares if they talked about the Palmares language. This was clearly the opinion of Rodrigues (1945: 137). However, two present-day authors claim that there was a language specific to Palmares. One of them is Décio Freitas, who says that “they could not adopt one of the native languages of Africa without sacrificing their unity. However, they needed a common language. Thus, the Palmarean language was formed, a syncretic language in which the African component had a decisive role but which equally incorporated elements from Portuguese and from Tupi” (Freitas 1984: 41-42). Clóvis Moura also says that there was a “Palmares dialect” (Moura 1984: 44-47).
I would like to add a contemporary evidence, on which Freitas and Moura based themselves,  in favor of “Palmarean creole”. It is a passage by the then (1678) governor of Pernambuco, Francisco Gomes Freire. According to him, the Palmares inhabitants “speak a completely different language, a language of their own, sometimes similar to the Guinean or the Angolan, sometimes similar to the Portuguese and Tupi language, but it is different from all of these. It is a new language” (apud Freitas 1984: 42). Further evidence is the fact that the whites did not understand the Palmareans. When they sent messengers to Palmares to discuss peace treaties, they had to take interpreters with them (Freitas 1987: 42). Surely, this unintelligibility was not due to the use of African languages -- which probably did not exist there -- as the governor’s quotation suggests.
Some of the few examples of the Palmarean language I could pick up in the historical documents are  person names like Ganga Zumba (Great Lord), Zambi (War God), Osenga, Lucrécia, Matias Dambi, Magdalena, and several others, besides place names like Angola Janga (Great Angola), an epithet used by Palmareans as an alternative name for Palmares. This expression could be a hint at the morphosyntax of Palmarean: the word order inside the NP would be noun + adjective although the name “Ganga Zumba” would be a counter-example.
The above linguistic data are insufficient to prove the existence of a language specific to Palmares. However, the probability that it did exist is even higher than for other parts of Brazil. After all, besides the testimony of the governor -- whose aim was to destroy Palmares -- it was made up of a multilingual population (P), living in a common territory (T) isolated from the rest of the country for almost 70 years, and forming a society with clearly defined rules. 

5.2.Guinea-Bissau

In the previous subsection I mentioned two cases of language contact in a T: one in which there is a certain probability that a creole may have existed (Brazil as a whole) and another in which it is almost certain that it did exist. In the present subsection, I will discuss a case in which a creole effectively arose, namely, in the Guinean coast of Africa. The creole that emerged there (henceforth Kriol) is one of the first creole languages to emerge out of the colonization of Africa, America and Asia by Europeans. I will specifically discuss the community of Cachéu. Geba, Bolama and Bissau are also important in this regard, but their history is less documented than Cachéu’s.
Cachéu was one of the first forts the Portuguese established on the Guinean coast. According to Valentim Fernandes, around 1505/1510 there were markets where the indigenous people traded their goods among themselves and with the Portuguese. Almada (1946) says that by 1590, the Portuguese had their quarter in Cachéu, separated from that of the blacks (p. 45). Later documents inform that the village (praça) of Cachéu consisted of three quarters. They were Vila Quente (Warm Village), Vila Fria (Cold Village) and Calaca. Besides that, there was the Rua do Taco (Taco street), a “feitoria” (factory), a “casa forte” (armory) and an “almazem” (storehouse).
Regarding the population (P) of Cachéu, around 1590 it was circa 1,500 people. In terms of concentric circles, there were, at the center, the Portuguese, most of them lançados (see Couto 1993). Many of the Portuguese had African wives or concubines -- dubbed “tangomas” by Almada -- with whom they had several children. These mixed-blood children were called “filhos da terra”, i.e., sons of the land. Around this first nucleus there were the grumettoes, helpers to the lançados, at first in the navigation of the several rivers of Guinea -- the original Portuguese word “grumete” designates a subordinate sailor. Around this second circle, there were several Africans who did not have direct contact with the Portuguese, although at least indirectly they dealt with them.
The small community composed of  “lançados”, “tangomas”, “filhos da terra” and grumettoes  was the originator of Kriol. In this connection, Valentim Fernandes recorded several words which were current there around 1510. Among them are jidiu (Jewish, dealer), bufri (buffalo?), kau (place), chon (ground, soil, land), kuskus (cuscus), and several others. Talking about the blacks of a neighboring village, Almda says that they “are cognizant of our language” (Almada 1946: 48). About this passage, the Guinean creolist Benjamim Pinto Bull said: “I think that by our language a kind of Portuguese creole should be understood” (Bull 1989: 71).
Around 1684, Francisco de Lemos Coelho recorded the first use of the word “crioulo” to designate the language of Cachéu (Coelho 1953: 153). This, incidentally, is a strong argument in favor of the hypothesis that the word “creole” originated in Portuguese, not in Spanish, as some French creolists seem to imply. In 1696, the first Kriol phrase was recorded, namely Agora mi sta sabi, i.e., “Now I am OK” (Portuense 1696).
In view of what has been said, it is no coincidence that creole speakers in Guinea-Bissau not only say that Kriol originated in Cachéu but also that they call it kriol fundu. Literally, this means “deep creole”. However, the meaning attached to the expression is rahter “true creole”, “pure creole”, “original creole”, etc. In other words, although the use of Kriol is widespread throughout Guinea-Bissau today, Guineans have a feeling that its origin is linked to a specific place, namely Cachéu.

6. Nativization or communalization?

We have already seen (section 5) that the concept of creolization as nativization of a prior pidgin is problematic. In the present section, I would like to explore these two concepts in more detail. Let me begin with the concept of pidgin. We have already seen that “pidgin” is frequently defined as “an auxiliary language that arises when speakers of several mutually unintelligible languages are in close contact; by definition, it has no native speakers” (Bickerton 1984: 173). Let us look at three phenomena that have been called pidgin, namely Tok Pisin (Todd 1990: 1-5), Lingua Franca (Cifoletti 1989) and Russenorsk (Hall 1966, Jahr 1989).
Lingua Franca was supposedly a pidgin based mainly on French, Italian, Spanish and North African varieties of Arabic, which was used around the Mediterrenean since at least the Middle Ages. However, when we look at the texts attributed to it, we can see that sometimes they look more like Spanish (si cane dezir doler cabeça, tener febre no poder trabajar ni saber como curar, a Fé de Dio abrusar vivo), sometimes more like French (quand moi gagner drahem, moi achetir mukere), and more frequently like Italian (mi star contento mirar per ti) (Schuchardt 1909: 457-458). That is, aren’t these utterances simply collections of SLA phrases uttered by Arabic speakers trying to speak one of these languages, or vice-versa? In fact, there seems to be no continuity, or transmission of an underlying grammar from generation to generation. Further, there is no specific place (T) where the community of speakers (P) of Lingua Franca (L) would be located, despite some claims that it would be very widespread in Algiers.
Russenorsk seems even more problematic, and this is due not only to the fact that only two PL were involved, and that both were at roughly the same level of power and/or prestige. What is more problematic is the fact that the encounters between Russians and Norwegians along the northern coast of Norway were seasonal and ephemeral. This means that there was no continuity in its use. Perhaps even more than in the case of Lingua Franca, the texts attributed to Russenorsk would be nothing more than sporadic utterances proffered by Russians and by Norwegians trying to speak each other’s language. When the Russians went back to their country, there remained no speakers of Russenorsk in Norway, nor did the Russians take Russenorsk with them to Russia.
If we consider Lingua Franca and Russenorsk pidgins, then we are not allowed to say that Tok Pisin is also a pidgin, even if we use Hymes’ distinction between “unstable pidgin” and “stabilized pidgin” (Hymes 1971). They are two completely different phenomena. Tok Pisin is the first and/or native language for many Papua New Guineans. In fact, it is a creolized language. If we use the same term to refer to such different phenomena, something seems to be wrong with it. 
Despite what has just been said, I think that Russenorsk and Lingua Franca are interesting in that they lead us to a more fundamental concept, namely that of language. Coming back to the phrases of these two “pidgins”, we could ask: Is it legitimate to call a collection of phrases a language? One could answer by stating that the philosopher Ludwig Wittgenstein said that “Die Gesamtheit der Sätze ist die Sprache” (the totality of propositions is language) (Wittgenstein 1966: 34, 35). However, many linguists would say that language consists first and foremost of a lexicon and a grammar. Others would add the texts that lexicon and grammar allow users to produce (Hjelmslev 1961: 15-29, Lyons 1996: 11-17). That is, the production of texts presupposes a language system (lexicon and grammar). Do Lingua Franca and Russenorsk have a lexicon and a grammar of their own? I doubt that they do, in spite of the fact that Slobin (1979: 43-46) and Jahr (1996) have presented some grammatical features of the latter. It should be noted that Jahr used the term “grammar of Russenorsk” in quotation marks (p. 110-114). A Gedankenexperiment would suffice to show that a mere set of utterances is not enough to prove that there is a language, even if Slobin and Jahr were right. Let us suppose that somebody collects a set of utterances proffered by monolingual speakers of Korean trying to speak English: 20 of them produced by speaker A in Miami in 1960, 30 by speaker B in Washington in 1970, 15 by speaker C in New York in 1990, and so on. A linguist who does not know the circumstances in which they were collected -- in disparate territories and in discontinuous time -- could arrive at some regularities which he could call the grammar of a “Koringlish” pidgin. The same could be said of utterances produced by children at a certain age. Are we allowed to speak of a “Childenglish”? If the answer is affirmative, we could ask where (T) are “Koringlish” and “Childenglish”  spoken.
The unavoidable consequence of what has been said above is that the concept of pidgin is too problematic to be considered the prerequisite for the emergence of a creole. I think that we could continue using “pidgin”, but referring to something that is not necessarily a potential prior creole, as admitted by Bakker (1996), Thomason (1997), and Mufwene (1997), among several others. However, it is necessary to define precisely what it means. If we exclude “pidgin” from the process of creole grammar formation, it is probable that what remains to be called “pidgin” would be what has been called “jargon”. If this is true, one of the terms is unnecessary. 
In summary, referring specifically to Russenorsk, we could say that if pidgin is a language, Russenorsk is not a pidgin; if Russenorsk is a pidgin, pidgin is not a language. This is true at least if we think of  language as consisting of a lexicon, a grammar, and the texts they allow the users (P) of the language to produce.    
Regarding the notion of nativization, it is also problematic in another sense. Departing from Gilbert’s (1986: 17) distinction between individual nativization and social nativization, we could ask: how many children of the emerging community must acquire the pidgin as their first language to make up a creole? According to Bickerton, one child would be enough (Bickerton 1991: 27-29), thus apparently accepting Gilbert’s individual nativization. However, we could ask him: “To whom would this solitary child talk?” (See Singler 1992).
Since both “pidgin” and “nativization” are problematic, it is unavoidable to think of another alternative to explain the emergence of creole languages. One alternative is already espoused by, inter alia, Albert Valdman (1977: 108), Hancock (1980: 64-65), Mufwene (1989: 75), Singler (1992), and Winford (1997). According to Baker (1990, 1997), when peoples speaking mutually unintelligible languages come together in a T, they create a common language of their own, which maay become the primary language of the new community. This means that a creole language emerges together with the emergence of a community. Therefore, instead of nativization, we could call the process of creole grammar formation communalization (communautarisation, in French). Moreover, since in order for a language to become the primary language of a community, there must be, in the first place, a place or territory where this community can be located -- the opinion of Samarin quoted above notwithstanding --, we could also speak of creolization as territorialization of a new language (see Labrie 1996).

7. Concluding remarks

The territory is the first foundation of a community. Thus, it would be a surprise if it did not play any role in the language spoken in the community. When peoples speaking mutually unintelligible languages have to share the same T, their individuals need to communicate with one another, first of all, in order to establish some rules for peaceful cohabitation or of good neighborhood. One of the first things they have to agree upon is which part of T is to be used by whom. Therefore, it is not astonishing that one of the first lexical items of the emerging community to appear is a name for T itself -- Cassidy (1971) gives some suggestions as to the probable first lexical items to emerge in such a situation.

This idea is so simple and obvious that one could question whether it is worth being discussed at all. In light of this, I would like to end by quoting one of the champions of the asocial view of language. According to him, "it is important to learn to be surprised by simple things -- for example, by the fact that bodies fall down, not up, and that they fall at a certain rate; that if pushed, they move on a flat surface in a straight line, not a circle; and so on". These simple ideas may lead us to “a surprising discovery, though the facts are entirely obvious to us” (CHOMSKY 1988: 43).

Note

*The ideas presented in this paper are part of a wider research project supported by CNPq, the Brazilian Council for the Development of Science and Technology, grant n. 201322/87-0. It represents my transition from creole studies to ecolinguistics. In it the fundamental environment of language was suggested for the firs time (1998). 

References

Alleyne, Mervyn. 1971. Acculturation and the cultural matrix of creolization. In Hymes (ed.): 169-186.
Almada, André Álvares de. 1946. Tratado breve dos rios da Guiné. Lisboa (1st ed., 1594).
Arends, J.  and P. Muysken & N. Smith (eds). 1996. Pidgins and creoles: an introduction. Amsterdam: John Benjamins.
Baker, Philip. 1990. Off target? Journal of pidgin and creole languages 5.1:107-119.
_______ 1997. Directionality in pidginization and creolization. Arthur Spears and Donald Winford, eds., 91-109.
Bakker, Peter. 1996. Pidgins.  J. Arends, P. Muysken and N. Smith, eds., 25-39.
Bakker, P. &  P. Muysken. 1996. Mixed languages and language intertwining. J. Arends, P. Muysken and N. Smith, eds., 41-52.
Bickerton, Derek. 1984. The language bioprogram hypothesis. The behavioral and brain sciences 7:173-221.
_______. 1988. Creole languages and the bioprogram. In Linguistics. The Cambridge survey. F. J. Newmeyer, ed., 268-284, Cambridge: Cambridge University Press.
_______. 1991. On the supposed gradualness of creole development. JPCL 6.1: 25-58.
Bollée, Annegret. 1977. Remarques sur la genèse des parlers franco-créoles de l’Océan Indien. In Meisel (ed.): 137-149.
Broch, Olof. 1927. Russenorsk. Archiv für slawische Philologie 41:209-262.
Bull, Benjamim Pinto. 1989. O crioulo da Guiné-Bissau: Filosofia e sabedoria. Lisboa: ICALP.
Cassidy, Frederick G. 1971. Tracing the pidgin element in Jamaican creole (with notes on method and the nature of pidgin vocabularies). In Hymes (ed.): 203-221.
Chomsky, Noam. 1988. Language and problems of knowledge: The Managua lectures. Cambridge: The MIT Press.
Cifoletti, Guido. 1989. La lingua franca mediterranea. Padova: Unipress.
Couto, Hildo Honório do. 1992. A linguagem do quilombo de Palmares. Papia 2.1:97-103.
_______ 1993. The genesis of Portuguese creole in Africa. In Atlantic meets Pacific. F. Byrne and J. Holm, eds., 381-389. Amsterdam: John Benjamins.
_______ 1994. O crioulo português da GuinJ-Bissau. Hamburg: Helmut Buske Verlag.
_______ to appear. On the question of the (prior) creolization in Brazil. MagnusHuber and M. Parkvall, eds., Spreading the word: Papers on the issue of diffusion of Atlantic creoles. London: Westminster University Press (Crelistics Series vol. 5).
Eibl-Eibesfeldt, Irinäus & W. Wickler. 1968. Ethology. International encyclopedia of the social sciences 5:186-193.
Fernandes, Valentim. 1951. Description de la côte occidentale d’Afrique (Sénégal au Cap de Monte, Archipels). Bissau: Centro de Estudos da Guiné Portuguesa (written around 1510).
Freitas, Décio. 1984. Palmares: a guerra dos escravos. Porto Alegre: Mercado Aberto, 5th ed.
Gilbert, Glenn. 1986. The Language Bioprogram Hypothesis: Déjà vu? P. Muysken and N. Smith, eds., 15-23.
Hall, Edward T. 1968. Proxemics. Current anthropology 9.2/3:.83-95.
Hall, Robert A. Jr. 1966. Pidgin and creole languages. Ithaca: Cornell University Press.
Halliday, M. A. K. 1976. Anti-languages. American anthropologist 78.3:570-584.
Hancock, Ian.; 1980. Lexical expansion in creole languages. Theoretical orientations in creole studies. A. Valdman and A. Highfield, eds., 63-88. New York: Academic Press.
_______. 1986. The domestic hypothesis, diffusion and componentiality. An account of Atlantic Anglophone creole origins. P. Muysken and N. Smith eds., 71-102.
Haugen, Einar. 1972. The ecology of language. The ecology of language, 325-339. Stanford: Stanford University Press: 325-339.
_______. 1979. Language ecology and the case of Faroese. In Linguistic and literary studies in honor of Archibald A. Hill IV. M. A. Jazayery et al., eds., 243-257. The Hague: Mouton. 
Hjelmslev, Louis. 1961. Prolegomena to a theory of language. Madison: The University of Wisconsin
Press.
Holm, John. 1988. Pidgins and creoles I. Cambridge: Cambridge University Press. 
_______ 1994. A semicrioulizaHno do portuguLs vern<culo do Brasil: EvidLncia de contacto nas expresstes idiomáticas. Papia 3.2:51-61.
Hymes, Dell. 1971. “Introduction”. Pidginization and creolization of languages. Cambridge: Cambridge University Press.
Jahr, Ernst H. 1996. On the pidgin status of Russenorsk. In Language contact in the Arctic: Northern pidgins and contact languages. E. H. Jahr and I. Broch, eds., 107-122. Berlin: Mouton de Gruyter.
Labrie, Normand. 1996. Territorialité. In Kontaktlinguistik. H. Goebel, P. H. Nolde, Z. Starv & W. Wölck, eds., 210-218. Berlin: Walter de Gruyter.
Lefebvre, Claire & J. Lumsden. 1989. Les languages créoles et la théorie linguistique. Canadian journal of linguistics 34.3:249-272.
Lyons, John. 1996. On competence and performance and related notions. In Performance and Competence in second language acquisition, Gillian Brown, Gillian, K. Malmkjaer, and J. Williams, eds., 11-32. Cambridge: Cambridge University Press.
Mackey, William F. 1979. Toward an ecology of language contact. In Sociolinguistic studies in language contact: Methods and cases. W. F. Makey, and J. Ornstein, eds., 453-459. The Hague: Mouton.
_______ 1980. The ecology of language shift. In Sprachkontakt und Sprachkonflikt. P. Nelde, ed., 35-41. Wiesbaden: Franz Steiner Verlag.
MacLuhan, Marshall and Q. Fiore. 1967. The medium is the massage. New York: Bantam Books.
Meisel, Jürgen M. (ed.) Langues en contact - pidgins - creoloes - Languafes in contact. Tübingen: Gunter Narr.
Montesquieu, Charles de Secondat, Baron of. 1899. The spirit of laws vol. I. New York: colonial Press.
Morgan, Lweis H. 1878. Ancient society. New York: Henry Holt and Company.
Moura, Cl\vis. 1987. Quilombo: a resistLncia ao escravismo. São Paulo: Editora Ática.
Mufwene, Salikoko. 1997. Jargons, pidgins, creoles, and koines: What are they? A. K. Spears and D. Winford, eds., 35-70.
Muysken, Peter and N. Smith (eds.) Substrate versus universals in creole genesis. Amsterdam: John Benjamins.
Preziosi, Donald. 1979. The semiotics of the built environment. Bloomington: Indiana University Press.
Reinecke, John E. 1937. Marginal languages: A sociological survey of the creole languages and trade jargons. Ph. D. dissertation, Yale University. Ann Arbor: University Microfilms International.
Rodrigues, José H. 1983. A vit\ria da lRngua portuguesa no Brasil colonial. Humanidades I.4:.21-41.
Rodrigues, Nina. 1945. Os africanos no Brasil. São Paulo: Companhia Editora Nacional, 3rd. Ed., chap. III.
Sapir, Edward. 1963. Language and environment. Selected writings, 89-103. Berkeley: University
Saussure, F. 1989. Course in general linguistics. La Salle, Il.: Open Court.
Schuchardt, Hugo. 1909. Die Lingua Franca. Zeitschrift für romanische Philologie XXX:441-461.
Sebeok, Thomas A. 1972. Perspectives in zoosemiotics. The Hague: Mouton
Silva, Leonardo Dantas. 1988. Alguns documentos para a hist\ria da escravidno. Recife: Editora Massangana.
Singler, John V. 1992. Rejoinder: Nativization and pidgin/creole genesis: A reply to Bickerton. JPCL 7.2:319-333.
Slobin, Dan. 1979. Psycholinguistics. Glenview, Il.: Scott, Foresman and Co. 2^nd ed.
Spears, Arthur K. & Donald Winford, eds., The structure and status of pidgins and creoles. Amsterdam: John Benjamins.
Taylor, Ralph B. 1988. Human territorial functioning. Cambridge: University of Cambridge Press.
Thomason, Sarah G. 1997. A typology of contact languages. A. K. Spears and D. Winford, eds., 71-88.
_______ and T. Kaufman. 1988. Language contact, creolization, and genetic linguistics. Berkeley: University of California Press.
Todd, Loreto. 1990. Pidgins and creoles. London: Routledge.
Valdman, Albert. 1977. Créalisation sans pidgin: le système des déterminants du nom dans les parlers franco-créoles. In Meisel (ed.): 105-136.
Wiesinger, Peter. 1973. Die deutschen Sprachinseln in Mitteleuropa. Lexicon der germanistischen Linguistik, 367-377. Thbingen: Max Niemeyer Verlag.
Winford, Donald. 1997. Column: Creole formation in the context of contact linguistics. JPCL 12.1:131-151.
Wittgenstein, Ludwig. 1966. Tractatus logico-philosophicus. London: Routledge & Kegan Paul.

[This paper was presented at the symposium on “Pidgin and creole languages in the 21st century”, Society for Pidgin and Creole languages, New York, January 1998].